Synthesis of Vane's prostaglandin X, 6, 9. alpha.-oxido-9. alpha., 15

Mar 1, 1977 - Prostaglandin synthesis. 16. The three-component coupling synthesis of prostaglandins. M. Suzuki , A. Yanagisawa , Ryori. Noyori. Journa...
0 downloads 0 Views 408KB Size
2006 00-

5a, X = 0

11

b, X = N C 0 , M e

Me 9

I 0

II

p

Me 6a, R =

OH;X = 0

b, R = OH; X = NC0,Me c, R = H;X = 0 d, R = H; X = N C 0 , M e

NCHO Me 7

F

Me 12

in terms of a stepwise mechanism involving a zwitterion (10 or l l ) ,not by a concerted mechanism.20 Thus the zwitterions (10, 11) are intercepted by the nucleophiles at low temperatures to give the hydroperoxides (3, 6) or rearrange to the dioxetanes (8,9) at ordinary temperature.22According to the MIND013 calculations, the zwitterion, an initial intermediate in enamine-singlet oxygen reaction, has been predicted to undergo rearrangement to a dioxetane with a relatively high activation energy compared to that for other processes such as rearrangement to a perepoxide.6a If so, it seems very likely that the lifetime of the zwitterions (10, 11) will be longer at lower temperature, permitting the trapping reactions more efficiently. The product ratio (6/7) is also solvent dependent. Polar solvents appear to increase the ratio of the dioxetane mode products (7) to the trapping reaction at least at 20 OC (Table I), although the solvent effect is still obscure. It is known that polar solvents increase the ratio of dioxetane formation to ene r e a ~ t i o n . ~ ' a . ~ s ~ 3 In order to get the spectroscopic evidence for the initial intermediate, we carried out the photooxygenation of 5a at -70 OC in an NMR cell. The NMR spectrum (-70 "C) of the reaction mixture in CD30D or CDCl3 had only the resonances of 6a. Neither zwitterion 11 nor dioxetane 9 could be detected at the t e m p e r a t ~ r eThe . ~ ~ spectroscopic studies at -70 OC provided no direct evidence in support of the zwitterions; however, we believe that the results described here may represent chemical evidence for the intermediacy of the zwitterionic peroxides.

Acknowledgment. This work was supported by a Grantin-Aid for Scientific Research from the Ministry of Education of Japan and the Japan Society for the Promotion of Science. References and Notes (1) Photoinduced Reactions. 97. (2) T. Matsuura and I. Saito in "Photochemistry of Heterocyclic Compounds", 0. Buchardt, Ed., Wiley, New York, N.Y., 1976, p 458. (3) (a).C. S.Fwte and J. W.P., Lin. TetrahedronLeft., 3267 (1968);(b) J. Huber. ibid., 3271 (1968); (c) T. Matsuura and I. Saito, ibid., 3273 (1968); (d) H. H. Wasserman and J. L. Ives, J. Am. Chem. Soc., 98,7868 (1976). (4) Recently, certain enamine dioxetanes have been isolated: (a) C. S. Foote, A. A. Dzakpasu, and J. W.-P.. Lin, Tetrahedron Left., 1247 (1975); (b) H. H. Wasserman and S.Terao, bid., 1735 (1975). (5) (a) D. R. Kearns, Chem. Rev., 71, 395 (1971); (b) P. D. Bartlett and A. P. Schaap, J. Am. Chem. SOC., 92,3223 (1970).

Journal of the American Chemical Society

(6) (a) M. J. S. Dewar and W. Thiel, J. Am. Chem. SOC., 97,3978 (1975); (b) K. Yamaguchi, T. Fueno, and H. Hukutome, Chem. Phys. Left., 22, 466 11973). (7) There are several examples of photooxygenations which have been explained in t e r m of zwitterions.2,3c% . e also (a) H H. Wasserman, Ann. N.Y Aced. Sci., 171, 108 (1970); (b) T. Matsuura and I. Saito, Tetrahedron, 25, 549 (1969); (c) I. SaRo, M. Imuta, and T. Matsuura. Chem. Left., 1173, 1197 (1972); (d) K. Orito. R. H. Manske, and R. Rodrigo, J. Am. Chem. Soc.,96, 1944 (1974). (8)(a) I. Saito, M. Imuta, S. Matsugo. and T. Matsuura, J. Am. Chem. Soc.,97, 7191 (1975); (b) I. Saito, M. Imuta, S. Matsugo. H. Yamamoto. andT. Matsuura. Synthesis, 255 (1976). (9) Irradiation was made with a tungsten-bromine lamp through an aqueous CuC12-CaC12 filter solution. (10) The hydroperoxide 3a readily decomposed in methanol with r l l Zof ca. 15 min at 30 OC to give a complex mixture of products including 2 (30%) am! polymeric materials. (11) All new compounds gave satisfactory elemental analyses and mass spectral data. (12) Viscous oil: starch-Kl test positive; UV (EtOH) 245, 294 nm; NMR (CDCI3) 8 1.58 (s, 3 H, Me), 2.90 (s. 3 H, NMe), 3.58 (s, 3 H. OMe). 4.45 (s, 1 H. NCHO), 6.30-7.40 (m, 4 H. arom H). 9.70 (s. OOH). (13) Bp 110 O C / 1 mmHg; UV (EtOH) 241 (log f 3.57). 285 nm (log 6 3.20); NMR (CDC13) 8 1.60 (s, 3 , Me), 3.18 (s, 3 H. NMe). 3.23 (s. 3 H, OMe), 3.35 (s, 3 H, OMe), 4.28 (s, 1 H, NCHO). 6.50-7.10 (m. 4 H. arom H). (14) Viscous oil: starch-KI test positive; UV (EtOH) 247, 295 nm; NMR (CDCl3) 8 1.30(t.3H9J=7Hr,CH2CH3). 1.55(~.3H.Me).2.88(~.3H.NMe).3.80 (4, 2 H. J = 7 Hz, CH&H3), 4.50 (s, 1 H, NCHO). 6.30-7.40 (m, 4 H, arom H), 9.65 (s, OOH). (15) Brief refluxing or standing (1 h) at room temperature of the solutions of 6a and 6b gave 6c and 6d, res ectively, in quantitative yield. (16) Recently, Nakagawa et al.' have reported the formation of 6b, 6d, and 7b in the photooxygenation of 5b in pyridine-methanol. The spectral data of 6b, 6d, and 7b were identical with those obtained by the authors. We are indebted to Professors T. Hino and M. Nakagawa for disclosure of their results prior to publication. (17) M. Nakagawa. H. Okajima. and T. Hino. J. Am. Chem. SOC..in press. (18) K. R. Kopecky. J. E. Filby, C. Mumford. P. A. Lockwood, and J-Y. Ding, Can. J. Chem.. 53, 1104 (1975). and references therein. (19) The possibility that 3 and 6 are formed from the dioxetane 8 and 9, respectively, cannot be ruled out completely, since the chemistry of indole dioxetanes has never been known. However, the temperaturedependency and the solvent effect on the formation of 6 are not satisfactorily ex lained by the dioxetane mechanism, whereas the MIND013 calculations& have predicted that polar solvents increase the ratio of the rearrangement of zwitterion (11) to dioxetane (9) in accordance wlth the experimental results. (20) Perepoxide such as 12 might also be proposed to explain the formation of 6a,b. While perepoxides have been proposed to rearrange to ene products andlor dioxetanes." " there is no precedent in which perepoxides have been considered to react with alcohols or amines. Note that the compounos (1, 5) having allylic hydrogens do not yield the ene products. (21) (a) N. M. Hasty and D . R . Kearns, J. Am. Chem. Soc.. 95,3380 (1973); (b) A. P. Schaap and G. R. Faler, ibid.. 95, 3381 (1973): (c) L. M. Stephenson, D. E. McClure. and P. K. Sysak. ibid., 95, 7888 (1973); (d) W. Ando, K. Watanabe. J. Suzuki, and T. Migita, ibid., 98, 6766 (1974). (22) Analogy for this process has been suggested in the reaction of hydroperoxides of pyrroles and imidazoles with base: (a) G. Rio, A. Ranjon. 0. Pouchot, and MJ. Scholl, Bull. SOC.Chim. f r . , 1667 (1969); (b) E H. White and M. J. C. Harding, Photochem. fhotobiol.. 4, 1129 (1965). (23) P. D. Bartlett, G. D, Mendenhall, and A. P. Schaap. Ann. N.Y. Acad. Sci.. 171, 79 (1970). (24) The NMR spectrum (-70 OC) of the reaction mixture resulting from the photooxygenation of 1 (CD30D. -70 O C ) also showed the presence of 3c as a sole product.

Isao Saito,* Mitsuru Imuta, Yoshiyuki Takahashi Seiichi Matsugo, Teruo Matsuura* Department of Synthetic Chemisrr), Faculty of Engineering, Kyoto Unicersity K ~ O I606, O Japan Receiced October 26. I976

Synthesis of Vane's Prostaglandin X, 6,9a-Oxido9a,15a-dihydroxyprosta-(Z)5,(E)13-dienoic Acid Sir: Vane and co-workers have recently obtained evidence for the formation of a new and remarkably active prostaglandin, termed PGX, from the prostaglandin endoperoxides PGG2 or PGH2 and microsomal fractions of certain tissues, especially aorta, arterial wall, and fundus of stomach.'Y2 Vane's PGX inhibits platelet aggregatioh as do PGEl and PGD2, but is several times more potent; it also causes relaxation of arterial smooth muscle. Although no structure was proposed for PGX,

/ 99:6 / March 16, 1977

2007

E"Z

1 *

HO

L -

THPO

OTHP

L c O'H

5

H

1

1

OH

,.3

2 u

Hoot--$: H RO L

-

4, R = THP

c

5R 6 H

l

l

l

l

5 , R = THP

7, R = H

,.6. R = H

d

0

HO %c5H11

HO

L HO

8

?cooH C

5

H HO

9

N

the genesis from the PG-endoperoxides (e.g., PGH2, l), the sensitivity to acid in aqueous solutions (rapidly increasing below pH 7), and its probable intermediacy in the formation of 6-oxo-PGF1,,~-~ all suggest that PGX is an internal enol ether of the latter, most likely a 6,9-enol ether 2, which can arise as shown. This possibility has now been confirmed by an unambiguous synthesis of PGX from PGF2, of natural configuration which also permits the assignment of the Z geometry to the 5,6-double bond of PGX as in 2. Reaction of the 1 l115-bistetrahydropyranylether of prostaglandin F2,4 (3) in THF-chloroform (25 mL/g of 3) with 1.1 equiv of N-bromosuccinimide at 23 "C for 1 h afforded the diastereomeric bromo ethers 4 and 5.5 Although these ethers were not readily separable by thin layer chromatography (TLC), depyranylation (acetic acid-water-tetrahydrofuran 3:l: 1 at 45 "C for 4 h) afforded the easily separable dihydroxy bromo ethers 6 and 7 in a ratio of ca. 3:1 (81% yield overall from 3; observed Rfvalues on silica gel TLC plates with benzene-dioxane-acetic acid 20:10:1 as solvent, 0.23 for 6 and 0.28 for 7).5The NMR and infrared spectra of 6 and 7 clearly indicate the absence of the cis-5,6-olefinic unit and the retention of the trans-] 3,14-double bond. Treatment of the major bromo ether 6 with excess potassium tert-butoxide in tert-butyl alcohol at 45 "C for 1.5 h to effect elimination of hydrogen bromide, concentration, rapid extraction of product with ether from a pH 5 aqueous layer cooled to 0 "C, and treatment with diazomethane afforded the acid sensitive methyl ester of 2.5,6In contrast the stereoisomeric bromo ether 7 was recovered virtually unchanged after exposure to potassium tert- butoxide under the conditions outlined above. These results indicate that the proton attached to C-6 in the bromo ethers 6 and 7 are exo and endo (i.e., less sterically hindered and more hindered), respectively, relative to the bicyclic nucleus, and together with the well-known trans addition pathway for bromo ether formation allow designation of the stereochemistry of 6 and 7. Further, the trans-coplanar course

of E2 elimination from 6 (which clearly would be followed here) must produce the Z geometry of the 5,6-double bond as shown in formula 2. Thus, the prostanoid 2 is readily available from 3 by an unambiguous and stereocontrolled synthetic route. Independent evidence for structure 2 was obtained by the extremely facile and clean hydrolysis of the methyl ester of 2 (in THF-0.01 M hydrochloric acid 3:1 at 23 "C for 10 min) to a more polar substance of Rf 0.17 in ether-acetone (3:1), which was characterized as 6-keto-PGF1, methyl ester by conversion to the known5 U-benzyloxime Samples of 2 were obtained for bioassay as the pyrrolidine salt by prompt treatment of the cold ethereal extract (described above) with 2-3 equiv of pyrrolidine, rapid concentration